Incidence of hyperoxia and related in-hospital mortality in critically ill patients: a retrospective data analysis

Abstract

Background: Mechanical ventilation with oxygen is life-saving, however, may result in hyperoxia. The aim was to analyse the incidence and duration of hyperoxia burden and related in-hospital mortality in critically ill patients.

Methods: Patients of all ages admitted to intensive care units (ICUs) and with mechanical ventilation for at least seven consecutive days were included in this single centre retrospective medical record audit. The main outcome measure was time-weighted arterial partial pressure of oxygen (PaO₂ ) over 7 days. Logistic regression for association with in-hospital mortality and propensity score matching was performed.

Results: In total, 20,889 arterial blood gases of 419 patients were analysed. Time-weighted mean PaO₂ was 14.0 ± 2.4 kPa. Time-weighted mean FiO₂ was 49.2 ± 12.1%. Seventy-six (18.1%) patients showed continuous hyperoxia exposure, defined as time-weighted mean PaO₂ > 16 kPa. Duration of hyperoxia, hypoxia (PaO₂ < 8 kPa) and normoxia (PaO₂ 8-16 kPa) were 37.9 ± 31.0 h (23.7%), 4.9 ± 9.5 h (3.1%), and 116.8 ± 29.6 h (73.2%). Hyperoxia occurred especially at low to moderate FiO₂ in patients of first and second age quartiles (1-57 years) with smaller SAPS2 score. In-hospital mortality of patients with hyperoxia (32.9%) or normoxia did not differ (35.9%; P = 0.691). Conditional logistic regression showed no association between hyperoxia and in-hospital mortality (OR 1.46; 95%CI 0.72-2.96; P = 0.29).

Conclusion: Substantial hyperoxia burden was observed in ICU patients. Young patients with less comorbidities showed hyperoxic episodes more often, especially with lower FiO₂ . Hyperoxia during 7 days of mechanical ventilation did not correlate to increased in-hospital mortality.